Neural networks to identify video encoding artifacts

Abstract

Apparatuses, systems, and techniques to perform neural networks. In at least one embodiment, one or more neural networks are used to identify one or more video encoding artifacts based, at least in part, on a plurality of different video encoding settings.

What is claimed is: 
     
         1 . A processor comprising:
 one or more circuits to use one or more neural networks to identify one or more video encoding artifacts based, at least in part, on a plurality of different video encoding settings.   
     
     
         2 . The processor of  claim 1 , wherein the plurality of different video encoding settings are used to generate a plurality of sets of images usable to train the one or more neural networks to identify the one or more video encoding artifacts. 
     
     
         3 . The processor of  claim 1 , wherein at least one of the plurality of different video encoding settings is generated randomly. 
     
     
         4 . The processor of  claim 1 , wherein the plurality of different video encoding settings are generated based, at least in part, on a statistical distribution over a range of video encoding settings. 
     
     
         5 . The processor of  claim 1 , wherein the plurality of different video encoding settings are generated based, at least in part, on a subset of a range of video encoding setting values. 
     
     
         6 . The processor of  claim 1 , wherein the plurality of different video encoding settings are selected based, at least in part, on a video encoding codec. 
     
     
         7 . The processor of  claim 1 , wherein the video encoding artifacts are identified based, at least in part, on a set of high-quality image frames and a set of encoded image frames. 
     
     
         8 . A computer-implemented method comprising:
 using one or more neural networks to identify one or more video encoding artifacts based, at least in part, on a plurality of different video encoding settings.   
     
     
         9 . The computer-implemented method of  claim 8 , wherein at least one of the plurality of different video encoding settings is used to generate a set of encoded image frames based, at least in part, on a set of high-quality image frames. 
     
     
         10 . The computer-implemented method of  claim 8 , wherein at least one of the plurality of different video encoding settings comprises a video encoding setting value generated randomly. 
     
     
         11 . The computer-implemented method of  claim 8 , wherein the plurality of different video encoding settings are generated randomly based, at least in part, on a normal distribution over a range of video encoding settings. 
     
     
         12 . The computer-implemented method of  claim 8 , wherein the plurality of different video encoding settings are generated randomly based, at least in part, on a thresholded subset of a range of video encoding setting values. 
     
     
         13 . The computer-implemented method of  claim 8 , further comprising:
 selecting a video encoding codec; and   generating the plurality of different video encoding settings based, at least in part, on the video encoding codec.   
     
     
         14 . The computer-implemented method of  claim 8 , further comprising:
 selecting a video encoding setting of the plurality of video encoding settings;   generating a set of encoded image frames using the selected video encoding setting; and   identifying the one or more video encoding artifacts based, at least in part, on the set of encoded image frames.   
     
     
         15 . A computer system comprising:
 one or more processors and memory storing executable instructions that, if performed by the one or more processors, use one or more neural networks to identify one or more video encoding artifacts based, at least in part, on a plurality of different video encoding settings.   
     
     
         16 . The computer system of  claim 15 , wherein at least one of the plurality of different video encoding settings is used to generate a set of encoded image frames. 
     
     
         17 . The computer system of  claim 15 , wherein at least one of the plurality of different video encoding settings comprises a video encoding setting value generated randomly. 
     
     
         18 . The computer system of  claim 15 , wherein the plurality of different video encoding settings are generated based, at least in part, on a statistical distribution over a range of video encoding settings values. 
     
     
         19 . The computer system of  claim 15 , wherein the plurality of different video encoding settings are generated based, at least in part, on a subset of a range of video encoding setting values. 
     
     
         20 . The computer system of  claim 15 , wherein the video encoding artifacts are identified based, at least in part, on one or more sets of high-quality image frames and one or more corresponding sets of encoded image frames.